There’s a moment I remember clearly from about six months into learning card magic.
I’d been watching performance videos online to study technique and presentation. Stumbled across a video of a sixteen-year-old demonstrating a flourish sequence — not an effect, just a visual display of card handling ability. The fluency was extraordinary. Hands that had been doing this for years, probably since childhood. The cards moved like water.
I was thirty-something. Six months into this. And I sat in a hotel room in Innsbruck holding a deck of cards thinking: I will never have that.
Not “I won’t have it yet.” I genuinely thought: I will never have that kind of embodied fluency. The window has closed. Adult hands don’t learn the way childhood hands learn. I have started too late.
I still think there’s a partial truth in that. There are things that become automatic at a neurological level when you practice them in childhood that require more conscious effort when you learn them as an adult. I will never have exactly what that sixteen-year-old had.
But David Epstein’s discussion of “desirable difficulties” in Range helped me understand something important: the slower, harder, more conscious process of adult learning often produces deeper knowledge — not despite the difficulty but because of it.
What Desirable Difficulties Are
The research on learning consistently shows a paradox: the learning methods that feel most effective in the moment are often the ones that produce the shallowest long-term retention and transfer. The methods that feel frustrating, slow, and ineffective in the moment often produce the deepest learning.
These counterproductive-feeling methods are called “desirable difficulties” by Robert Bjork, one of the psychologists who has studied them most extensively.
The most studied desirable difficulties are:
Spaced repetition — spreading practice over time rather than massing it. Massed practice (practicing everything in one long session) produces fast apparent progress. The material feels learned quickly. But it fades quickly too. Spaced practice (returning to material across many sessions separated by gaps) is slower and feels less satisfying, because you keep forgetting material between sessions and having to re-learn it. But that re-learning process builds durable retention.
Interleaving — mixing different types of material in a single practice session rather than blocking (practicing one thing until you’ve got it, then moving to the next). Blocked practice feels more productive because you’re visibly improving at each thing in sequence. Interleaved practice is frustrating because you keep jumping between things before any one feels solid. But interleaved practice builds the ability to identify which approach applies to which situation — a discrimination skill that blocked practice doesn’t build.
Testing/retrieval — deliberately trying to recall material rather than re-studying it. Re-studying is easy and feels productive. Trying to retrieve material you can’t quite remember is frustrating. But the act of retrieval attempt — even failed retrieval — substantially strengthens retention.
Why Adult Learning Is Actually Different, and Better
Here’s what I missed in that Innsbruck hotel room moment.
The sixteen-year-old’s learning was fast and produced high automaticity because he had a childhood brain, enormous practice volume from a young age, and no competing cognitive framework to create interference. He learned like water flowing into a container.
My learning was slow because I was an adult with an existing, complex cognitive architecture. Every new skill had to be integrated into a mind that already had thousands of existing patterns, beliefs, frameworks, and memories. New material doesn’t flow in cleanly — it has to find its relationship with everything already there.
This integration process is slower. It feels harder. It often feels like failing.
But the integration, precisely because it’s active and effortful, produces something the sixteen-year-old’s fast learning doesn’t produce: explicit understanding of how the new skill relates to everything else I know.
He can perform the flourish with extraordinary automaticity. Ask him why the timing works, how it creates the visual impression it does, what cognitive principles underlie the audience’s response — he may or may not be able to articulate it. The knowledge is embodied and automatic but possibly not deeply understood in a transferable way.
My slower learning forces understanding. I can’t rely on pure automaticity, so I have to actually understand what I’m doing. The understanding is effortful, but it’s durable and transferable in a way that pure automaticity isn’t.
The Specific Desirable Difficulty I Benefit From
The most important desirable difficulty in my learning is the gap between sessions.
Because I’m not a full-time performer, I practice in intervals. Hotel room sessions. Early mornings before keynotes. Occasional weekends. I’m not doing this eight hours a day. I’m doing it in concentrated bursts separated by gaps.
This drives me crazy from a pure progress standpoint. I’ll have a practice session where something starts to feel right, then come back three days later and it’s degraded. I have to rebuild it. That re-building feels like going backward.
But the research says: that re-building is where the deep learning happens. The material I can rebuild after three days of forgetting is material I actually know. The material that’s only solid in the flow of a single long practice session is material I merely processed, not material I learned.
The gaps are working in my favor. The frustration of having to re-establish things is a signal that the desirable difficulty is doing its job.
The Transfer Problem
There’s a related issue that the desirable difficulties research illuminates: transfer.
Transfer is the ability to apply what you’ve learned in one context to a different context. It’s the thing you actually want from learning — the ability to use what you know flexibly, adaptively, in new situations.
Learning in a massed, comfortable, uninterrupted way tends to produce context-specific knowledge. You know how to do the thing in the conditions where you practiced it. Change the conditions and the knowledge fails to transfer.
Learning through desirable difficulties — spacing, interleaving, testing — tends to produce more flexible knowledge. Because you’ve had to retrieve and apply the knowledge in varied conditions, the underlying structure is more accessible. You can apply it when the conditions change.
This maps directly to the wicked learning environment problem I described in an earlier post. Performance contexts vary enormously. An effect that’s been practiced to automaticity in hotel room conditions may not transfer to a formal stage setting, or to an outdoor corporate event, or to a small close-up context in a bar.
If my practice has been through desirable difficulties — spacing, interleaving, varying conditions — I have a better chance of transfer. If my practice has been comfortable, massed, and consistent-condition, I have an automaticity that’s brittle in the face of variation.
What I Do Differently Because of This
Practically, understanding desirable difficulties changed my practice in a few specific ways.
I stopped trying to consolidate material in single sessions. Instead of practicing something until it feels solid, I practice it briefly, move to something else, and come back. The incompleteness of each session’s progress is now a feature rather than a bug.
I deliberately practice in varied conditions. Hotel rooms are my primary studio, but they’re not identical hotel rooms. Different furniture arrangements. Different lighting. Standing versus seated. Full run-throughs versus isolated components. The variation prevents the context-specific learning that makes transfer fragile.
And I maintain a review practice. Old material comes back into rotation deliberately, not just when I’m preparing for a specific performance. The gaps between touching a piece of material grow over time as it becomes more durable, and return to being frequent if I notice degradation.
This is nothing exotic. It’s standard spaced repetition logic. But building it into an irregular practice schedule — which is what my life requires — means it has to be deliberate rather than emergent.
The teenager with the extraordinary flourishes learned something real and valuable. I don’t want to diminish that.
But I learned differently, more slowly, with more frustration, with more conscious struggle. And that slower, more difficult path built a different kind of knowledge — one that transfers, that articulates, that connects to other things I know.
Feeling behind wasn’t a sign that I was failing to learn. It was a sign that I was learning in the way that adults actually learn things well.
Which is slow. Which is hard. Which feels like going backward sometimes.
And which, in the end, goes deep.